FCS Phase II Results Paper 3 - Detailed Technology ...



AERONAUTICAL COMMUNICATIONS PANEL (ACP)

Working Group C – 11th meeting

Brussels, Belgium

18 – 20 September 2006

Agenda Item 3 Update on Future Communication Infrastructure Activities

FCS Phase II Results Paper 3 - Detailed Technology Investigations

Prepared by: Glen Dyer, Tricia Gilbert / ITT Industries

James Budinger / NASA Glenn Research Center

SUMMARY

The Future Communication Study (FCS) Phase II detailed investigations described herein were conducted primarily in response to feedback received on FCS Phase I technology investigation results that raised concern about technology performance in the L-band aeronautical channel environment; ground infrastructure cost for an L-Band air-ground (A/G) communications solution; satellite solution performance; and technology performance in the C-Band aeronautical surface environment.

To address the received feedback and to support in-depth evaluation of technologies emerging from the Phase II screening process, several detailed study activities were conducted as part of FCS Phase II technology evaluation. These included investigations specific to the L-Band aeronautical channel including definition of a channel model that could be used for common characterization of waveform performance in the A/G channel; definition of a framework for specifying the infrastructure costs associated with an L-Band system; and analysis of the performance of recommended technologies with the common channel model and their potential to interfere with incumbent users of the band. Also investigated was satellite system modeling (with a focus on availability analysis) and C-Band performance modeling (to evaluate technology performance on the airport surface).

This paper provides an overview of the FCS Phase II detailed investigations and provides recommendations to the group specific to the described results.

1. Background

As part of the Future Communication Study (FCS) cooperative research program, candidate technologies supporting the long-term aeronautical mobile communication operating concept are being evaluated. After Phase I of the FCS technology investigation, Technology Pre-Screening, feedback from several stakeholders was received. This feedback indicated that there was concern about technology performance in the L-band aeronautical channel environment; ground infrastructure cost for an L-band A/G solution; satellite solution performance; and technology performance in the C-Band aeronautical surface environment.

With this feedback in mind, Phase II of the FCS technology investigation, Technology Screening, commenced. The Technology Screening process evaluated an inventory of more than 50 technologies consisting of both commercial standards/systems as well as custom aeronautical standards. These technologies were organized into technology families characterized by similarities in user requirements, services offered, and reference and physical architectures. The screening of the technologies considered technology performance with regard to data loading capability, technology communication range, and ability to use protected spectrum. Specific threshold values were associated to these metrics traceable to requirements in the Concept of Operations and Communication Requirements for the Future Radio System (COCR). A result of the screening was the identification of the most promising technology candidates to bring forward as candidate technologies for the Future Radio System (FRS).

As a result of the technology screening process described in a companion paper, eight technologies have been identified as candidates for a general aeronautical communication solution for the FRS (also called a continental solution because the solution applies to all continental flight domains including airport, terminal and en route). In addition, some additional technologies have been identified as the best performers in the context of specific flight domains that have a unique environment and may warrant separate technology considerations (i.e. oceanic and airport domains). A summary of the results is provided in Table 1 below.

Table 1: Recommended Technologies from Technology Screening

| |NASA/ITT Screened Technologies |

|Continental Solution |W-CDMA |

| |APCO P34 |

| |L-band E-TDMA |

| |LDL |

| |B-VHF (at L-band) |

| |Link 16 |

| |Inmarsat SBB |

| |Custom Satellite Solution |

|Oceanic Domain |Inmarsat SBB |

| |Custom Satellite System |

|Airport Domain |IEEE 802.16 |

To further understand the technologies emerging from the screening process and to address specific feedback received on the Phase I technology investigation results, Phase II of the technology investigation included a set of focused and in-depth analyses. The topics of these studies were organized into five major areas including:

• L-Band Technology Performance

• L-Band Interference

• L-Band Technology Cost for Ground Infrastructure

• Satellite Technology Availability Performance

• C-Band Technology Performance

It is the intent of this paper to provide an overview of the detailed study results specific to the topics above.

1. L-Band Technology Performance

Consideration is being given to the use of L-Band (960-1024 MHz) to employ the next generation aeronautical communication system. Several screened technologies are being considered for this band, including P34, LDL, WCDMA, B-VHF (at L-Band), and L-Band E-TDMA. Upon review of the technology definitions and developed concepts of use, technology considerations warranting in-depth analysis were identified. The focus of Phase II in-depth technology investigations was specific to P34 and LDL, two technologies that scored well during Phase I technology investigations. The selected analysis topic(s) for each technology was made based on the need to assess those components of the technology that provide the most challenge for application of the technology as a viable solution for aeronautical communication as well as received feedback on Phase I results. Selected analysis topics for the candidate L-Band technologies include:

• P34: Protocol model developed in OPNET to assess P34 net entry and data transfer performance and BER performance in the L-Band channel

• LDL: Bit-Error-Rate (BER) performance in the L-Band channel

As both P34 and LDL analysis topics include assessment of their performance in the L-band channel, a first step in the L-band analysis work was modeling of this channel. A literature search revealed that while many channel models exist for the terrestrial channel in close proximity to L-Band, there had been no previous activity to develop a channel model that characterizes the L-Band Air/Ground (A/G) channel for radiocommunications. As most standardization bodies consider it a best practice to test candidate waveform designs against carefully crafted channel models that are representative of the intended user environment, a channel model was developed that could be used for common characterization of communications waveform performance in this A/G channel.

Characterization of the Delay Spread and the Doppler Power Spectrum is essential for generating a useful model for waveform simulation and evaluation of candidate FRS technologies in L-Band. In order to form estimates of the delay spread and associated statistics, a ray-tracing simulation was developed. This simulation models both diffuse and specular reflections from the Earth’s surface. The specific channel modeled was one associated with a worst case scenario, specifically modeling of the A/G channel over mountainous terrain. This terrain, in the en route case, has the potential to provide long multipath delays that either limit the data rate to be transmitted or require special techniques to achieve acceptable performance. The context of the developed model for evaluation of the A/G channel is shown in Figure 1.

Figure 1: L-Band Channel Model Context

[pic]

The specific implementation of the model was for mountainous terrain in Aspen Colorado. A topographic map of this terrain is provided in Figure 2.

Figure 2: Aspen Colorado Mountain Terrain Used for L-Band A/G Modeling

[pic]

The developed simulation used a method of concentric oblate spheroids to model multipath contributions. Using the transmitter and receiver as focal points, a series of oblate spheroids was generated in the three-dimensional simulation space. The first oblate spheroid in the series was generated so that it just barely intersected the underlying terrain. The semi-minor axis of each successive oblate spheroid was increased by a fixed increment so that the spheroids intersected more and more of the underlying terrain as they were stepped through. The desired product was the set of points on the terrain that were intersected by the oblate spheroids. When plotted, each set of intersection points appears as a distorted annulus approximating the cross section of the spheroid when sliced by the Earth’s surface. Each set of intersection points is mutually exclusive from any other set because any intersection point can only be accounted for once. Each set of intersection points contributes to multipath for a particular delay. Figure 3 illustrates the method of concentric oblate spheroids used to model multipath contributions.

Figure 3: Two Concentric Oblate Spheroids Intersecting the Underlying Terrain

[pic]

The contour of terrain trapped between two successive spheroids was used to calculate multipath dispersion for a particular time delay. Each contour consisted of a set of terrain points that represented potential scatterers. Ray tracing was used to determine specular and diffuse multipath. The detailed methodology utilized to identify multipath components is shown in Figure 4.

Figure 4: L-Band Channel Modeling– Identification of Multipath Components

[pic]

Implementing the methodology above and employing data reduction and analysis techniques, the mean RMS delay spread was calculated to be 1.4 µs. It is instructive to consider representative technologies at this point since the technology data rate will drive channel model parameter estimation. A rule of thumb frequently applied is that if the mean RMS delay spread is at least one tenth of the symbol duration, then the channel is frequency-selective. In order to illustrate this, two technologies that scored well during the FCS Pre-Screening were considered: LDL and P34. Table 2 shows the corresponding data rates and symbol durations for LDL and P34.

Table 2: Data Rates of LDL and P34

|Waveform |Data Rate |Symbol Duration |1/10th of the Symbol |

| | | |Duration |

| |[pic] |[pic] |[pic] |

|LDL |62.5 kbps |16 µs |1.6 µs |

|P34 |4.8 kbps* |208.3 µs |20.83 µs |

* P34 is an OFDM system. The tabulated data rate is per carrier and is the symbol rate. Overall P34 data rates range from 76.8 – 691.2 kbps

Using our rule of thumb, P34 should undergo flat fading and LDL presents a borderline case because the mean RMS delay spread is very close to one tenth of the symbol duration. It is important to note that frequency-selective channel models differ in structure from flat fading channel models. For this reason it was decided to develop a frequency-nonselective fading model for P34 and a frequency-selective fading model for LDL. An example of the resulting, more complex model for LDL is shown in Table 3 below.

Table 3: LDL Channel Model Parameters

|Tap # |Delay (µs) |Power (lin) |Power (dB) |Fading Process |Doppler Category |

|1 |0 |1 |0 |Rician |Jakes |

|2 |1.6 |0.0359 |-14.5 |Rayleigh |Jakes |

|3 |3.2 |0.0451 |-13.5 |Rayleigh |Jakes |

|4 |4.8 |0.0689 |-11.6 |Rayleigh |Jakes |

|5 |6.4 |0.0815 |-10.9 |Rayleigh |Jakes |

|6 |8.0 |0.0594 |-12.2 |Rayleigh |Jakes |

|7 |9.6 |0.0766 |-11.2 |Rayleigh |Jakes |

One of the primary results reported is the simulated RMS delay spread. It should be noted that this delay spread can be modeled as a function of the average distance from the transmitter, with increasing delay spreads reported for increasing distances. Specifically, a generalized model, using the method cited in Greenstein, has the form:

where,

• d is the distance in km

• σ0 is the median value of the RMS delay spread at d = 1 km

• ε is an exponent that lies between 0.5-1.0, based on the terrain type

• A is a lognormal variate

To determine the parameters that are appropriate for a generalized L-Band A/G model in mountainous terrain, RMS delay spreads were predicted for a reference distance of 1 km as well as for the previously mentioned values at 64.37 km (40 miles). The two predicted values that resulted from the simulation work are:

(RMS(1 km) = 0.1 μs

(RMS(64.37 km) = 1.4 μs

Fitting the Greenstein model to the reference data provides a generalized expression for RMS delay spread, which is found to be:

Having defined a channel model to support FRS technology performance analysis, the focus is now turned to evaluation of two specific candidate technologies, P34 and LDL. As noted above, P34 in-depth analysis included assessment of P34 net entry and data transfer performance and BER performance in the L-Band channel. To assess net entry and data transfer performance, a simulation was developed. The simulation work included development of an operational scenario, communication nodes and communication links. The selected scenario for evaluation was the NAS Super Sector, as defined in the COCR. In this scenario, one fixed station node was used to model the ground station and 95 mobile nodes were used to model aircrafts. The defined communication link for this model implemented the P34 Scalable Adaptive Modulation (SAM) air interface. This included 50 kHz channels and QPSK modulation (providing 76.8 kbps). This is the lowest defined P34 data rate, but should be satisfactory for “closing the link” for the sector size defined in the COCR. A depiction of the simulation context for this analysis is shown in Figure 5.

Figure 5: P34 Simulation Context

[pic]

Using this context, a P34 configuration that was simulated was fixed-network equipment (FNE) to mobile radio (MR). The MR to MR and repeater modes were not simulated. The simulated configuration aligns with the P34 concept of use for aeronautical application defined during Phase I of the FCS technology investigation. The custom OPNET development included modeling of the P34 PHY, MAC, LLC and SN layers, as illustrated in Figure 6.

Figure 6: Modeled Elements of P34

[pic]

Simulation model results are shown in Figure 7. These figures show the response time of the P34 simulation to the offered load for each of the transmitted message. Note that the sub-network latencies over P34 protocols (SNDCP, LLC CP, LLC UP, MAC) meet COCR latency requirements. Specifically, although there are some startup outliers, 95% of delay measurements are under 0.7 seconds.

Figure 7: P34 OPNET Modeling Results

[pic]

In addition to simulation of P34 to evaluate net entry and data transfer performance, P34 performance in the defined L-Band A/G channel was also evaluated. As part of this effort, P34 transmitter and receiver models were generated. Specifically, the P34 SAM physical layer interface was modeled by developing a custom application using C code. The transmitter was implemented as detailed in the specification for the 50 kHz channel using QPSK modulation. The receiver implementation was tested against published results for standardized channel models. Figure 8 shows known results for P34 in Additive White Gaussian Noise (AWGN) and the HT200 channel model (on the left side) and simulated performance in these channels (on the right side).

Figure 8: Validation of P34 Receiver Model

[pic]

It was unclear from initial modeling results (shown above) if satisfactory performance was being achieved in the mobile fading channels. Specifically, the need to know the meaning of a raw BER of 3*10-3, or what effective error rate this translates to after coding, was identified. P34 SAM uses a system of concatenated Hamming codes as shown in Figure 9. The rate ½ coding was simulated by concatenating two Hamming coders and a block interleaver.

Figure 9: P34 SAM Coding Scheme

[pic]

Coding gain for P34 was calculated as shown in Figure 10. Here, a 3*10-3 raw BER is approximately equal to 10-5 coded BER.

Figure 10: P34 Coding Performance Simulation Results

[pic]

The developed P34 transmitter and receiver models were combined with a model of the expected L-Band channel based on analysis work previously described. Specifically, a two tap channel model was simulated where Tap 1 was modeled as Rician, with a K-factor of 18 dB, unity gain, and Jakes Doppler Spectrum; and Tap 2 was modeled as Rayleigh, with a 4.8 μs delay, -18 dB average energy, and Jakes Doppler Spectrum. In this model, the mobile velocity was taken to be .88 mach. This is the maximum domestic airspeed given in the COCR based on Boeing 777 maximum speed. Additionally, in the model the P34 tuned frequency was taken to be 1024 MHz, with maximum Doppler shift of 1022 Hz.

Initial simulations indicate good performance can be achieved in the aeronautical channel, primarily a consequence of the strong line-of-sight component of the received signal (with K factors greater than four). Figure 11 shows initial results (note that initial results are still being validated).

Figure 11: P34 Predicted Performance in the L-Band Aeronautical Channel

[pic]

A second technology investigated for performance in the L-Band aeronautical channel was LDL. As with P34, LDL transmitter and receiver models were generated and the receiver model validated against known results. A plot of theoretical performance versus achieved simulation performance is shown in Figure 12. Here the theoretical curve is the performance of a binary Continuous Phase Frequency Shift Keying (CPFSK) signal with coherent detection using n = 5 and h = 0.715 (reference Proakis). The simulation model uses the same traceback length (n = 5) and modulation index (h = 0.715). Note that using a modulation of 0.715 minimizes the probability of error for binary CPFSK (reference Schonhoff 1976).

Figure 12: Validation of LDL Simulation Model

[pic]

As with the P34 analysis, coding gain for the LDL model was also investigated. First, however, an adjustment in the modulation index was made. A modulation index of 0.715 was required to validate the model with published results, however LDL calls for a modulation index of 0.6. This change in the modulation index pushes the BER curve out approximately 1 dB. The Reed-Solomon (RS) coding defined for LDL can then be accounted for. Specifically, Reed-Solomon (72, 62) code provides a coding gain of 3 to 4 dB in the expected region of operation. In order for the RS code to provide a substantial coding gain, the raw BER must be less than 10-2 and ideally it should be less than 2*10-3. LDL simulated BER performance for h=0.6 and RS coding is shown in Figure 13.

Figure 13: LDL BER Performance (h=0.6 and RS Coding)

[pic]

The final step in LDL performance analysis was simulation of LDL in the L-Band aeronautical channel environment. The LDL channel model is a conservative model that introduces an irreducible error floor to system performance. The plot shown in Figure 14 shows the system performance of LDL in the presence of both AWGN and the L-Band aeronautical channel model. Based on the results of this simulation, LDL will require channel equalization to mitigate the effects of the A/G aeronautical channel model in L-band.

Figure 14: LDL Predicted Performance in AWGN and the L-Band Aeronautical Channel

[pic]

2. L-Band Interference Analysis

Several candidate technologies for the Future Radio System have been considered in the context of operation in the aeronautical L-Band spectrum. This band, 960-1215 MHz, has a primary allocation for Aeronautical Radio Navigation Services (ARNS). There are current several system implementations that occupy the band. ICAO systems that use spectrum in this band include the Universal Access Transceiver (UAT); secondary surveillance radars (including ATCRBS, Mode A and C, and Mode S); and Distance Measuring Equipment (DME). A majority of the spectrum allocations for these systems are standardized by ICAO. There are, however, some exceptions such as DME allocations defined on a national basis between 962 and 977 MHz in the US.

As part of the consideration of new future communication system technology implementations in this band, the need to analyze the interference potential of proposed technologies to systems currently operating the aeronautical L-Band spectrum has been identified. A generic process for interference analysis would have the following elements:

• Describe the source of interference and the interference mechanism

o Description is usually in the form of power spectrum and time characteristics (e.g., transmit (Tx) power, Transmission bandwidth, duty cycle)

• Quantify the isolation between transmitter output and receiver input

o This isolation includes the effects of antenna gains, cabling losses and propagation

• Determine the ratio of undesired to desired signal power at the input of the receiver decision process (detector)

• Quantify receiver performance as a function of this D/U ratio, ascribe a required performance and assess compatibility

The last item noted above is the most difficult element of the process and was the focus of the interference simulation work defined for this study. Specifically, during consensus deliberations among FAA, NASA and ITT at the beginning of the study, two technologies were selected for detailed analysis, LDL and P34. At that time, it was determined that the compatibility of those two proposes systems with existing ICAO standardized civil aviation systems, namely UAT and Mode S, would be included in the detailed analysis.

Several interference results were collected for evaluation of the interfering impact of LDL and P34 on UAT performance. Figure 15 (left side) below provides a collection of BER curves for varying degrees of LDL interference into the UAT signal. This figure (right side) also provides a collection of BER curves for varying degrees of P34 interference into UAT signals. From the curves it would appear that a Carrier-to-Interferer (C/I) ratio of between 12 and 15 dB is required for minimum degradation to the UAT receiver. LDL has slightly better performance than P34 in terms of not interfering with UAT receiver.

Figure 15: Interference Modeling for UAT

[pic]

A second set of results was collected for evaluation of the interfering impact of LDL and P34 on Mode S performance. Specifically, a set of probability of correct preamble detection curves were collected (based on an algorithmic assumption to declare preamble detection with 94% [model 1] and 100% [model 2] correlation). These results are shown in Figure 16 below with results for preamble detection declared with 94% correlation on the left side and with 100% correlation on the right side of the figure.

Figure 16: Mode S Probability of Correct Preamble Detection Curves (P34 and LDL Interferers)

[pic]

Additionally for Mode S, a set of probability of false preamble detection curves was also collected. These results are shown in Figure 17.

Figure 17: Mode S Probability of False Preamble Detection Curves (P34 and LDL Interferers)

[pic]

The modeling results would seem to indicate that a C/I ratio of 15 dB or better is required to not substantially degrade the Mode-S preamble detection performance. The behavior of “false preamble detection” would appear to be somewhat worse than the behavior of “missed preamble detection”. As in the UAT case, the performance of LDL is better than that of P34 – that is to say, P34 acts as more of an interference source than LDL to both Mode S and UAT receivers. It should be noted that all simulations were made “on-tune”; actual deployment scenarios should be far off-tune, especially for the Mode S case (proposed band for the FRS is 960-1024 MHz, and the Mode S Extended Squitter equipment is at 1090 MHz). Additionally, measurements should be made that further characterize Mode S behavior as there are other metrics to investigate besides preamble detection. Finally, the preamble detection modeled here is hardly sophisticated, and better performance from actual equipment is predicted.

3. L-Band Technology Cost for Ground Infrastructure

One additional detailed analysis performed with regard to L-Band technologies (in general, rather than specific to one technology) was the evaluation of economic feasibility from the perspective of the ground infrastructure provider. This analysis was responsive to feedback received on the initial technology pre-screening results that indicated that due to cost constraints, an L-Band solution is only considered should VHF spectrum prove insufficient to provide total required data link capability. The L-Band business case analysis provided a first order of magnitude estimate of required investment for an L-Band aeronautical ground infrastructure. The technical approach for accomplishing this objective included:

• Through detailed analysis, develop a notional ground L-Band architecture that can meet Future Communication Infrastructure (FCI) requirements as defined in the Communications Operating Concept and Requirements for the Future Radio System (COCR) document for ATC communications

o Derive number of radio sites required for total US coverage

▪ Perform L-Band link budget analysis

• Develop L-Band Link budget spreadsheet and derive the parameters to close the link

• Excess Path Loss derivation

▪ Perform L-Band coverage analysis

▪ Derive radio site redundancy to meet system availability requirements

▪ Develop an architecture to meet availability required

• Determine if the business case can close

o Develop cost elements and estimates for initial development and O&M

o Determine required revenue flow to close business case

An overview of the technical approach work flow is shown on Figure 18.

Figure 18: Process for Determining Service Provider Cost

[pic]

While the first order of magnitude cost estimate yielded positive business case results, the important aspect of the study to bring forward to ACP was the framework of the analysis which can be considered a generic framework specifying infrastructure costs associated with an L-Band system. Along with the methodology shown above, the L-band cost modeling work employed several assumptions for consideration including:

• L-Band provides coverage to a large continental region (e.g. United States or Core Europe)

• Coverage is above FL180

• System Availability of Provision (Ap) meets COCR requirements for COCR Phase II en route services (sans Auto-Execute service)

• Cost elements considered include: Research and Development (including system design and engineering); Investment (including facilities and equipment); and Operations and Maintenance (including telecommunications, personnel and utility costs)

4. Satellite Technology Availability Performance

For the Satellite and Over Horizon technology family, two technology inventory candidates have emerged from the technology screening: Inmarsat SwiftBroadband (SBB) and Custom Satellite Solution. The Inmarsat SBB candidate differs from many of the other candidates considered in that it is an operational system with a defined service architecture and a defined set of service offerings. For this candidate, the ability to meet COCR performance requirements was selected as the focus of detailed analysis.

COCR performance requirements are specified for data capacity, latency, QoS and maximum number of users, but there are also availability and integrity requirements for aeronautical services. The performance of Inmarsat SBB with regard to capacity, latency, QoS, and number of users was evaluated along with other screened technologies as part of the preliminary iteration of detailed technology analysis. This being the case, the selected focus for the in-depth analysis was other COCR performance requirements, specifically availability performance. Availability was selected as it was considered as a potential shortfall of the satellite candidate solutions.

The Custom Satellite Solution is a technology concept that includes the fielding of a custom satellite or custom satellite payload specifically designed for aeronautical communications. This concept is being explored by several civil aviation authorities and related organizations. Japan has launched an aeronautical communication satellite and is exploring performance of next generation satellite systems. The FAA’s Global Communication, Navigation, Surveillance System (CGNSS) contract Phase I study explored the definition of a satellite architecture for providing aeronautical safety services. And finally, there are consortiums that are working to define aeronautical satellite specifications, such as the Satellite Data Link System (SDLS). Within this body of work, the need to accommodate all communication safety services with associated performance requirements has been considered. It has been found that to meet these requirements, a highly reliable, highly available architecture is required, such as the five satellite architecture proposed in the GCNSS study.

Here again, availability arises as an important issue. In order to provide required availability, a highly redundant custom satellite system architecture is needed. As this issue is similar to that noted above for Inmarsat, a separate study of availability for Custom Satellite Solutions was not performed. Rather, it was considered to be more instructive to estimate the availability of two existing, operational satellite systems, Inmarsat SBB and Iridium that provide services in protected aeronautical spectrum (AMS(R)S).

In summary, the focus of the detailed analysis for satellite technologies was availability performance. The performance of existing AMS(R)S systems (namely Inmarsat SBB and Iridium) was examined. Calculated availability of these architectures was contrasted with the calculated availability of a generic VHF terrestrial communication architecture, specifically a data communication architecture based on existing infrastructure.

The approach used for SATCOM availability modelling was the analysis model described in RTCA DO-270. This document defines an availability fault-tree to permit characterization and evaluation of multiple availability elements. The fault tree is organized into two major categories, system component failures and fault-free rare events. This model, shown in Figure 19, was useful for comparing architectures and was applied in this study.

Figure 19: SATCOM Availability Modeling Approach – Fault Tree

[pic]

A summary of availability modeling results is shown in Figure 20. For SATCOM systems, limiting factors for availability include satellite equipment failures and RF link effects (Inmarsat and Iridium), capacity overload (Iridium), and interference (Iridium). For the VHF terrestrial reference architecture, the limiting factors for availability include RF link events and capacity overload. Overall, the detailed evaluation of satellite communication systems (with a focus on provision of required availability) indicated that both Inmarsat SBB and Iridium would not meet availability requirements. Also, custom satellite solution designed to meet COCR availability requirements would, in fact, require a highly redundant and costly architecture. For these reasons, the satellite solutions are not considered viable solutions for the continental domain. This does not preclude their effective role in providing communication capability in remote and oceanic airspace.

Figure 20: Summary of Availability Modeling Results

[pic]

5. C-Band Technology Performance

The C-Band modeling activities were conducted to investigate the utility of an industry standard system in the airport surface environment. The system that was chosen for analysis was the IEEE 802.16e Metropolitan -Area Network (MAN) interface standard. The IEEE 802.16e standard (referred to as simply the 802.16e standard, or 802.16e henceforth) was chosen as it scored well during the initial phase (technology pre-screening) of the FCS technology investigations.

As the 802.16e standard supports a range of physical layers, prior to the modeling process, a specific physical layer needed to be selected. Of the possible candidates, better mobility performance is expected from OFDMA than OFDM, and the leading commercial 802.16 forum (the WiMAX Forum) has defined “Mobile” WiMAX profiles which are all expected to adopt the OFDMA physical layer. In this study, however, the OFDM physical layer was selected for analysis, as it seems that if good performance can be predicted for OFDM then by inference the OFDMA physical layer would also work well. Further, there are commercially available chipsets for the 802.16 OFDM physical layer currently available. Since a logical next step to this research would be prototype implementations and trials in the band, and noting that OFDM (due to the aforementioned chipset) is more amenable to prototype equipment development, this seemed to be a reasonable decision.

The process for modeling the 802.16e OFDM physical layer was to:

1. Select a modeling environment

2. Implement the 802.16e transmitter functions and validate against known test vectors

3. Implement additional transmitter functions required to simulate system and validate by inspection of output signal spectrum

4. Implement 802.16e receiver and validate end-to-end performance in an Additive White Gaussian Noise (AWGN) channel with known results

5. Implement a fading channel for which published results are available, implement the receiver channel equalization, and validate model performance

6. Introduce the airport surface channel model as defined by Ohio University research and published in “Wireless Channel Characterization in the 5 GHz Microwave Landing System Extension Band for Airport Surface Areas,” Final Project Report for NASA Glenn Research Center ACAST Project, Grant Number NNC04GB45G, May 2006.

Implementing the methodology defined above, 802.16e transmitter and receiver functions were modeled in the MATLAB Simulink® environment. A depiction of the transmitter model is provided in Figure 21 and a depiction of the receiver model is provided in Figure 22.

Figure 21: 802.16e Transmitter Model

[pic]

Figure 22: 802.16e Receiver Model

[pic]

The next step in the C-Band Modeling work was to validate the developed model. Specifically, the simulation was executed in an AWGN environment and corresponding results compared to published results. Good correlation was achieved as shown in Figure 23 below.

Figure 23: Validation of 802.16e Simulation Model

[pic]

Using a channel model adapted from a detailed model developed by Ohio University, the performance of 802.16e in the aeronautical airport environment was simulated as shown in Figure 24. Here performance was found to be quite good for most of the movement area (incorporating equalization techniques). While this technology has good potential applicability for this domain, additional analysis to look at additional technology features to enhance performance (e.g. Hybrid Automatic Repeat Request (HARQ), fast feedback channel and diversity sub-carrier permutations) is warranted.

Figure 24: 802.16e Simulation Results for the Aeronautical C-Band Surface Channel Model

[pic]

6. ACTION BY THE MEETING

The working group is invited to consider the technology investigation activities described in this paper and provide comments if desired.

It is recommended that the ACP Working Group consider the A/G channel model that is presented in this paper and adopt it for the evaluation of candidate technologies for the Future Radio System.

It is recommended that the ACP Working Group consider the cost modeling approach that is presented in this working paper and adopt it for the evaluation of candidate technologies for the Future Radio System.

-----------------------

[pic]

[pic]

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download